Burnt VRM?

[Fergutor]

Hi!
I just got an used Zotac GTX 970, this one https://www.zotac.com/tm/product/graphics_card/gtx-970 and is giving me the famous (I just found out...) "no-signal-crash" in every game that is so much common with those cards.
I know that nobody knows what causes these, and people give the most varied supposed fixes, things that supposedly and at least for a short time worked for them.
But this particular card has like a burned area in the back of the PCB just at the back of the VRMs. (I tried a lot of other possible fixes and I installed an Asus Strix GTX 970 that does not crash and works flawlessly, but that one uses 1 x 8 PCI-E cable while the Zotac uses 2 x 6 PCI-E, that's an important difference. The Asus is more factory overclocked than the Zotac). Also the last times I tested it, i felt a burning smell!
So look these pictures I took today, do you think its burnt? It's like that dirty part is burned (the guy talked much about hoe he take care of his parts but this says otherwise....)
One is a crop from one of those. I guess I could upload full sized ones if needed.


So, what do you think?
BYW I didn't check the temperature of the VRMs, simply because I didn't see at the moment if I could get acces to such readings (now whit the Asus I see that I can). I made many tests but not sure of any conclusion, for example used Afterburner to limit "power limit", underclocked the core, raised the fan speed and those only seemed to delay the crash (downclocking to around -500mhz I think it didn't crash but also didn't tested the games it crashed the most...my bad). Temperatures reach 80ºC but allowing the card to surpass that to almost 90º didn't caused crash as I was also underclocking...Wel the susual studd with this problem.
What do you think? Thanks and sorry for bad english.

 

[MrGenius]

Doesn't look burnt to me. All I see is dirt(dust).

 

[Fergutor]

Doesn't look burnt to me. All I see is dirt(dust).

Don't you think the colour there is suspicious?

 

[lexluthermiester]

I'd say those VRMs might be on their way out. The PCB looks mighty toasted. Might have got a bum card.

 

[natr0n]

Get a toothbrush and clean that area to see.

How are you supposed to examine something with shit smeared on glasses.

 

[MrGenius]

Don't you think the colour there is suspicious?

Not really. It just looks like it needs cleaned. And even if it is a little discolored from heat, that doesn't necessarily mean there's a problem. There might be, or there might not be. Some parts can run pretty hot without issues. If what you're showing me are the parts on the opposite side of the board there, they don't look like they've been overheated at all.

BTW, I'm not doubting you have a hardware issue(it sounds like you probably do). I'm just not sure if you've found it or not.

 

[Fergutor]

Get a toothbrush and clean that area to see.

How are you supposed to examine something with shit smeared on glasses.

Ah yeah, forgot to explain that: I'm going to take the card to the seller tomorrow.

Not really. It just looks like it needs cleaned. And even if it is a little discolored from heat, that doesn't necessarily mean there's a problem. There might be, or there might not be. Some parts can run pretty hot without issues. If what you're showing me are the parts on the opposite side of the board there, they don't look like they've been overheated at all.

BTW, I'm not doubting you have a hardware issue(it sounds like you probably do). I'm just not sure if you've found it or not.

Thanks.

I'd say those VRMs might be on their way out. The PCB looks mighty toasted. Might have got a bum card.

I think the same.

 

[Fergutor]

Well I grabbed a toothbrush and cleaned the area and ended up looking pretty good...

All the brown seemed to go away...
But the card was dying anyway...We tested it on a third machine and gave the same error (all three machines with good PSUs). So the guy had to return my money to me.

But still for information to someone else having this problems: I tested another card, of the same model and worked without problems. Also the same failing card seemed to work more stable with the provided molex to pcie adapters which is an interesting point, in fact I don't think it crashed once with those. In my machine, both with the pcie connectors as with the molex to pcie adapters, as in the 3rd guy's machine, the card made a very loud coild whine, while in the owner's machine the coil whine was a lot less noisy. This guy's PSU was a not modular silver 900w Cooler Master, mine is a semi modular Cooler Master 700w bronze SIlent Pro and the 3rd guy a not modular 650w, I think EVGA.

 

[eidairaman1]

Circle where you think the problem is.

 

[biffzinker]

All I'm seeing off is the capacitors I circled.

 

[lexluthermiester]

All I'm seeing off is the capacitors I circled.
View attachment 120687

Though discolored, they look ok.

 

[eidairaman1]

Though discolored, they look ok.

They are poor pictures without the items circled by the OP...

 

[Fergutor]

Circle where you think the problem is.

No that picture was aimed at natr0n and the others that commented on the aspect of that area before cleaning it.

They are poor pictures without the items circled by the OP...

I didn't circled them. By "poor pictures" you mean lacking information/context or the picture quality? It wasn't meant to obtain any diagnostic because as I said in the same comment I posted the photo, I returned the card, and the picture was meant to show how it looked without all that dirt on it to the guys that commented before. But if you are curious here you have a 100% size but at 50% quality (to reduce size):